Sept. 20, 1956
SPECTROSCOPIC STUDIES ON DYES: STRUCTURE OF IKDANTNKONES [CONTRIBUTION FROM THE
4599
QUARTERMASTER RESEARCH & DEVELOPMENT CENTER]
Spectroscopic Studies on Dyes.
111. The Structure of Indanthronesl
BY GEORGE M. WYMAN RECEIVED FEBRUARY 27, 1956 The infrared spectra of indanthrone, some of its halogenated derivatives and N-methylindanthrone fail to exhibit the expected -NH- stretching absorption band near 3.0 p . They show a strong band near 6.3 p, which is also shown by the corresponding anthraquinoneazines and which is probably characteristic of a -C=Nstretching vibration. On the basis of these observations a revised structure (111) is proposed for indanthrone. The seemingly anomalous chemical properties of indanthrone, such as its chemical stability, behavior in the vatting reaction, salt-formation with anhydrous strong bases, etc., are also readily explained on the basis of this structure.
Introduction The discovery of indanthrone, a bright blue dye of exceptional light and wash-fastness properties, a t the turn of the century2ushered in a new era in the development of synthetic dyes. Even now, more than 50 years later, indanthrone derivatives are still among the most sought after of the blue dyes that are available. The determination of the structure of indanthrone was the subject of a series of papers by Scholl and co-workers shortly after the discovery of this dyestuff. As a result of these studies they assigned structure I to indanthrone, and this 0
sxl I/
1
o
w
N ~
O I
+$ /&,Yo I1 0
I
formula has generally been accepted since. Although this structure seemed largely consistent with what was known about the chemical behavior of indanthrone a t the time, i t was somewhat dificult to reconcile i t with the blue color of the dye, since such a deeply colored substance might have been expected to possess a fully conjugated structure. Moreover, the azine (11), which results when indanthrone is oxidized, is a dull yellow in color even though it has a fully conjugated struc0
O
N
/I
(1) Presented hefore t h e Division of Organic Chemistry a t t h e 119th Meeting of t h e American Chemical Society, Dallas, Texas, April, 1956. ( 2 ) K. Bohn, German Patent 129.845 (1902). (3) (a) R . Scholl, Ber.. 36, 3410 (1903); (b) R. Scholl and H. Berblinger, i b i d . , 36, 3427 (1903); ( c ) R . Scholl, H.Berlinger and J. Mancfielrl, ihid , 4 0 , 320 (1907); ( d j R . jcholl, W. Steinkopf and A. Kdhacznik, i b t d . , 4 0 , 390 (1007).
ture. Aware of this apparent anomaly, Scholl attributed the deeper color of the dye t o the auxochromic effect of the -NH- groups upon the two quinone chromophores. 3a. It was this inconsistency between the structure and the color of this dye and of the corresponding azine that made it appear desirable to reopen this problem. Therefore, it was proposed to investigate indanthrone and some of its derivatives with the aid of spectrophotometric techniques that have only recently become available and to re-examine the data in the literature in the hope of reconciling these anomalies. Experimental Dyes.-Indanthrone and 3,3’-dichloroindanthrone were research samples obtained through the courtesy of Mr. P. Kronowitt of Ciba States Ltd. 3,3’-Dibromoindanthrone and the 1-methylamino-2-bromoanthraquinone which was used in the synthesis of N-methylindanthrone were kiiidly provided by Dr. 0. Stallmann of the du Pont Co. SMethylindanthrone and N,N’-dimethylindanthrone wcre prepared according to the procedure of Bradley and Leete.4 The methylated derivatives were identified by comparing their visible spectra in pyridine solutions with the spectra reported by these investigators. Oxidation of 1ndanthrones.-The anthraquinoneazines used in this investigation were prepared by the oxidation of the appropriate indanthrone with nitric acid, according to Scholl’s procedure.80 Deuteration of Indanthrone.-Twenty mg. of iiidaiithrone was added to 20 ml. of a 570 solution of XaOD in DzO contained in a 50-ml. erlenmeyer flask. This suspension was heated to 60-65” and sufficient sodium hydrosulfite added to effect reduction of the dye. The flask was thcn loosely stoppered and the deep blue solution maintained at this temperature for 5 minutes with periodic gentle swirling. The vat was reoxidized by blowing dry air through it for 5 minutes. The precipitate was filtered, washed with a small amount of water and dried a t 85”. Combustion of the product, followed by mass-spectrometric analysis of the water formed in the combustion indicated that the replacement of the active hydrogen atoms by deuterium was a t least 357, effective under these conditions. Measurements in the Infrared Region.-The samples (as mulls or as pellets in KBr) were measured by means of a Beckman IR-3 spectrophotometer, utilizing NaC1 or LiF optics, as required. The pellets used were 13 mm. in diameter and a pellet of pure KBr was used as the reference. In addition, where the entire infrared spectrum was desired, the pellets were also measured on a Baird Associates double-beam spectrophotometer equipped with a NaCl prism. Measurements in the Visible Region.-The Cary spectrophotometer (Model 11) was used for measurements in this region. Spectra were measured on solutions (against the solvent as reference) and on KBr pellets.6
Discussion The salient feature of the infrared spectra of the indanthrones that were studied in this work is the (4) W. Bradley and E. T e r t e , J . Cherir. S o r . , 21 17 (1931). ( 5 ) G , hr, T\’yman, .I O p t . S o r .-1rii , 4 6 , 91i.j (19.5;).
4600
GEORGEM. WYMAN
absence of any absorption band in the N-H stretching region (near 3.0 p ) . This wholly unexpected result was carefully verified by making several measurements using different techniques of sample preparation (e.g., mulls in Nujol or in a polytrifluorochloroethylene oil and as KBr pellets) and by observing that other compounds containing N-H groups adjacent in space to carbonyl groups, (e.g., 1-aminoanthraquinone and indigo) exhibited the typical N-H stretching band under the same conditions of measurement (cf. Tables I and 11) Although absorption bands that are due to the stretching of 0-H linkages have been known to become almost indistinguishable as a result of hydrogen-bonding,6 such weakening and broadening has never been reported for N-H bands.? T o the contrary, indigo exhibits a distinct N-H stretching band a t 3.05 p8 in the solid phase, even though the N-H group is involved in both intraand intermolecular hydrogen-bonding. Moreover, the partial replacement of the two active hydrogen atoms of indanthrone by deuterium (which would be expected to shift any N-H bands from the 3 p region, where they may have been obscured, to wave lengths longer than 4 p ) also failed to give rise to any new absorption bands. Consequently, the absence of the N-H stretching frequency presents a strong argument against the presence of the N-H groups in the indanthrone molecule.
TABLE I1 PRINCIPAL INFRARED ABSORPTION BANDSOF
In K B r
As Nujol mull
Iiidanthrone .. .. .. l\T-Methylindanthroiie .. .. .. 3,3’-Dichloroindanthrone .. .. .. 1-Aminoanthraquinoiie 2.92,3.02 2.92, 3 . 0 2 2.92,3.02 Indigo 3.06 3.06 3.06 5,5’,7,7’-Tetrabromoindigo 2 97 2.97 2.97
Another conspicuous feature of the infrared spectra of most of these compounds is the strong broad band near 6.3 p . I n indanthrone this band is almost as strong as the carbonyl band (cf. Fig. l), and it is also present in the spectra of the azines studied. Although this is the wave length range where aromatic ring vibrations are likely to occur, it is believed that these would not give rise to such an intense band as that shown by indanthrone. Instead, this strong band is very probably the C=N stretching frequency which would be expected to obscure the weaker aromatic ring vibrations that (6) (a) D. Hadzi, Abstracts of t h e X I V t h International Congress of Pure and Applied Chemistry, Zurich, 1955, No. 23, p. 14; (b) K. Nakamoto, M . hlargoshes and R. E. Rundle, THIS JOURNAL, 7 7 , 6480 (1955). (7) Cromwell, e l al., could not detect t h e N-H stretching frequency in the infrared spectra of some 8-amino-a,B-unsaturated ketones. However, they attributed this t o a shift of t h e N-H bands t o slightly longer wave 1en:ths (near 3.4 p ) where they would be obscured by t h e C-€I stretching frequencies (N. H. Cromwell, et al., THIS JOURNAL, 71, 3337 (1949)). (8) 1%’. I
